Re: [v6ops] Stateful SLAAC (draft-ietf-v6ops-unique-ipv6-prefix-per-host)

[Bob Hinden <[e-mail]>]

Ole,

> On Nov 13, 2017, at 10:42 AM, Ole Troan <[e-mail]> wrote:
> 
> Fernando,
> 
>>> <mailto:[e-mail]>> wrote:
>>> 
>>>   FWIW, I would agree with that if this were an issue of WG focus creep.
>>>   AFAICT, the issue appears to go much deeper, which means it's an Area
>>>   boundary issue if it is indeed a protocol extension.
>>> 
>>>   IMO, OPS stays in the lane of suggesting sets of *existing* protocol
>>>   parameters and features, or indicates where MAYs and SHOULDs can be
>>>   relaxed (or not) - all of this remains compliant with the protocol.
>>>   Changes to the protocol should not be considered operational decisions,
>>>   again IMO.
>>> 
>>> 
>>> Excuse me, but I really cannot fathom why we are saying that this draft
>>> defines a new protocol
>> 
>> Model SLAAC with a FSM for the client, and a FSM for the server. Now
>> apply this document. And look at the SLAAC router FSM.
>> 
>> Did it change (and quite a lot, actually)?  If yes, then you ahve
>> changed the protocol. If not, you didn't do the FSM properly.
> 
> Or do as I do in my implementation.
> Model each host as being on it's own point to point interface.
> Configure the IPv6 prefix on that interface. That configured state is exactly like what we have in classic SLAAC.
> 

Good point.  If the document had described this as something like creating virtual interfaces, we might not be having this discussion.

Bob